We compare three quasi-experimental approaches to estimating the returns to schooling in Australia: instrumenting schooling using month of birth, instrumenting schooling using changes in compulsory schooling laws, and comparing outcomes for twins. With annual pre-tax income as our measure of...